Your Resume Is Summarized Before Anyone Reads It

In partnership with

Yes, that Arnold Schwarzenegger.

He has a newsletter on beehiiv. So do Codie Sanchez, Jay Shetty, David Begnaud, Colin & Samir, and Joanna Stern.

They could publish just about anywhere. They chose beehiiv to build a direct connection with the people who want to hear from them most. And you can do the same, whether you’re world-famous or just at the start of your plan for world domination.

Get started for free.

Welcome to today's SCALIS EarlyCareers newsletter! 🚀

You submit into a pool of four hundred. In your head there is a picture of what happens next: a keyword parser, a match score, a ranked list. You have been optimizing against that picture for two years.

Last Monday, Northeastern published its AI and Society Survey on hiring. The researchers looked at 8,174 respondents who said they have a say in who gets hired. Fifty four percent said artificial intelligence is involved in reviewing resumes at their organization. In tech, that number reaches 83 percent. The report's own phrasing is the part worth sitting with: increasingly, software reads the resume before a person does.

Notice who answered. Not recruiting operations. Not the people who configure requisitions. People with a say in who gets hired, which for an internship is very often a team lead who got handed a packet and twenty minutes. Roughly half of them said AI made a difference in their decisions. Among that group, 26 percent said they are hiring fewer entry-level candidates, and 28 percent said they expect more output or higher-level work from the ones they do hire.

So the reader changed. A keyword parser and a general-purpose model fail you in completely different directions, and almost every piece of resume advice you have been given is written for the first one. Today is how to write for the second.

100+ coding prompts top engineers use to ship 5X faster

Claude Code, Codex, and Cursor are on every engineer's stack. Most still treat them like a search bar. Top engineers work from a system, these 100+ prompts are that system. Sign up for The Code and get the prompts free, plus a 5-minute daily newsletter to keep sharpening your edge.

Claim your free prompts

4,000+ BRAND NEW internship positions freshly posted today. View the full list below and start applying!

A parser skips you. A summarizer flattens you.

A parser's failure mode is omission. It cannot read your two-column layout, so a section vanishes and you sort low. The fix is mechanical: single column, standard headings, text-based file.

A summarizer's failure mode is different and worse. It reads everything you wrote and then compresses it, and compression keeps proper nouns, numbers, and named artifacts while quietly discarding adjectives, hedges, and anything that required the reader to infer. Nothing is omitted. You are just reduced to a version of yourself that sounds like everyone else.

Both failure modes are live right now, so keep the mechanical fixes. Then add the second job, which nobody is coaching you on.

Decide which three sentences you want to survive

Open your resume and ask the honest question: if someone compressed this to three sentences, what comes out? For most students the answer is "junior at a good school, relevant major, two leadership roles, decent GPA, interested in the field." That summary is true, generic, and indistinguishable from thirty others in the same batch.

Pick the three facts you want in that compression and make them structurally impossible to drop. Specific nouns in the top block rather than the bottom. Numbers attached to outcomes, because numbers survive compression and claims do not. The same through-line repeated in two places rather than five unrelated strengths listed once each.

Compare what compresses out of these two lines.

The first compresses to "marketing experience." The second compresses to the numbers, because that is what is left when everything else is squeezed out.

Shared facts cancel. One unshared fact decides.

When a manager drops a batch into a model, the question they type is almost always a version of "which of these stand out and why." That is a comparison prompt, and comparison prompts are brutal to anything you hold in common with your peers.

The same three courses everyone in your major took. The same club. The same case competition. The same GPA band. All of it is real, and all of it cancels in a comparison, because it appears on most of the other documents in the same window.

So audit your resume for exactly one thing nobody else in your graduating class could have written. A tool you built that someone else uses. A dataset you assembled. A number you moved that belongs to a real organization. You do not need five of these. You need one that cannot be canceled out.

Turn AI into Your Income Engine

Ready to transform artificial intelligence from a buzzword into your personal revenue generator?

HubSpot’s groundbreaking guide "200+ AI-Powered Income Ideas" is your gateway to financial innovation in the digital age.

Inside you'll discover:

• A curated collection of 200+ profitable opportunities spanning content creation, e-commerce, gaming, and emerging digital markets—each vetted for real-world potential

• Step-by-step implementation guides designed for beginners, making AI accessible regardless of your technical background

• Cutting-edge strategies aligned with current market trends, ensuring your ventures stay ahead of the curve

Download your guide today and unlock a future where artificial intelligence powers your success. Your next income stream is waiting.

Get Your Guide

Say how you used AI on your own schoolwork

Here is the most counterintuitive finding in the whole report, and students are getting it backwards almost universally.

Around 40 percent of hiring decision-makers said that using AI on every writing or coding assignment would be a point in a candidate's favor. Only 21 percent said it would discourage them from making an offer. The people deciding are roughly twice as likely to read your academic AI use as a plus than as a problem, and the survey found they are also the group most likely to expect AI to make their own work more productive.

You have been trained by syllabi to treat this as something to conceal. In the application, it is evidence. The version that works is specific about where the judgment was yours.

That sentence does something no model can fabricate on your behalf: it describes a real workflow with a real boundary in it.

You are being compared to the tool, not only to other students

Remember the 28 percent expecting higher-level work from entry-level hires. The quiet question behind an intern decision in 2026 is no longer only "is this student better than that student." It is also "is this student worth more than what I can already do without them."

Answer it on purpose. In interviews, and in any open field on the application, name the thing you do that does not get cheaper when the tools get better: you show up in a room, you own a mistake, you ask the question nobody wanted to ask, you verify the number before it leaves the building. Pair it with one example. Generic enthusiasm reads as the one thing on your resume a model could have generated.

Run the screen on yourself tonight

This takes eleven minutes and it is the closest you will get to seeing what the person on the other side sees.

Open a model, paste in the job posting, paste in your resume, and type the prompt a busy manager would actually type:

Read the three sentences cold. If they describe someone you would not interview, the problem is not the model. It is that you wrote a document that compresses into a stranger. Fix the top block, add the number, cut the adjective, and run it again.